Population delineation of barren-ground grizzly bears in the central Canadian Arctic

We identified spatially distinct populations of barren-ground grizzly bears (Ursus arctos) in a 235,000-km(2) study area northeast of Yellowknife, Northwest Territories, Canada. We tested for the presence of population clusters of bears using movement data obtained from satellite telemetry (n=55 female bear-years, n=42 male bear-years) and multivariate cluster analysis. We identified geographic range of female and male population clusters using the fixed-kernel range estimator with least-squares cross-validating to determine bandwidths, and defined population unit boundaries by referring to the 70% contours of population ranges. To validate population units, we required spatial clusters for female and male bears to be similar so distinctive female and male components could be contained within common population boundaries. Further, we needed population growth rates to be a result of intrinsic birth and death rates and not immigration or emigration rates. Thus, no more than 1 radiotracked animal of either sex (between 2.1% and 4.3% of a given population unit sample) could immigrate to or emigrate from an identified population unit annually. We obtained independent clustering solutions that grouped both female and male grizzly bears into 3 areas: the North Slave region, Bathurst Inlet region, and Kugluktuk region. Although female population ranges at the 70% contour level were completely contained within established population unit boundaries, male population ranges demonstrated overlap. Annual exchange rates were high (3.4-13% for females, 7-35% for males) among the 3 population groups. Bears in our study area should thus be managed as an open (continuous) population.